Probiotics as applied to humans are defined as live microorganisms which, when administered in adequate numbers, confer a health benefit on the host. The most frequently cited reasons for this probiotic activity include the production of anti-microbial substances such as bacteriocins and lactate and interference with Salmonella adhesion to the intestinal wall.
Infection with Salmonella results in millions of cases of human foodborne illness every year; the pathogen source varies, but many cases result from the consumption of contaminated porcine meat products (Swanenburg et al., 2001; Anonymous, 2002a). Awareness of food safety issues during all stages of pig meat production is thus vital, particularly with regard to reductions in levels of Salmonella contamination. Probiotics, as usually defined, are “live microorganisms which, when administered in adequate amounts, confer a health benefit on the host” (FAO/WHO, 2001). Probiotic properties have been ascribed to many microbial species, but those most commonly used are members of the lactic acid bacteria (LAB) group, particularly Lactobacillus and Bifidobacterium strains. Several studies have investigated the anti-Salmonella effects of potential probiotics using in vitro procedures, particularly growth medium and tissue culture (Drago et al., 1997; Fernandez et al., 2003).
A number of authors have linked the application of LAB probiotics with beneficial effects in models of gastrointestinal infection using small animals. Ogawa et al. (2001) reported that the use of Lb. casei Shirota reduced colonization levels and decreased the severity of diarrhoea in E. coli O157:H7-infected infant rabbits. Using mice, Johnson-Henry et al. (2004) have noted that a mixture of Lactobacillus strains reduces gastric inflammation and bacterial colonization in Helicobacter pylori-infected animals. Varied results have been reported with the use of Salmonella infection models. Pascual et al. (1999) noted complete exclusion by 21 days of Salmonella enteritidis using Lb. salivarius in chickens. Recently, La Ragione et al. (2004) observed no beneficial link between pre-treatment with Lb. johnsonii and Salmonella enteritidis faecal numbers or colonization of the chicken intestine. The same authors did however note that E. coli numbers were reduced in the small intestine, but not in the colon, caecum or faeces. They also claimed that the strain was very effective against Clostridium perfringens. Silva et al. (2004) observed improved survival for mice pre-treated with Bif. longum during Salmonella challenge, but no effect on numbers of the pathogen. They postulated that this may be due to a reduced inflammatory response mediated by the probiotic treatment, but not population antagonism.
Many studies investigating the effect of LAB probiotics on gastrointestinal infection in humans concentrate on antagonism of rotavirus infection in infants and Clostridium difficile infection in adults. Several reports have claimed beneficial effects for probiotics (particularly Lb. rhamnosus GG) in rotaviral infection of children; many of these are reviewed in Alvarez-Olmos and Oberhelman (2001). Among the initial publications in this area was the report by Isolauri et al. (1991) which demonstrated that treatment with Lb. rhamnosus GG reduced the duration of rotavirus diarrhoea in children. Reid et al. (2003) have reviewed data claiming a reduction in occurrence of Clostridium difficile diarrhoea in humans due to probiotic treatment with Lb. rhamnosus GG and Saccharomyces boulardii. 
With regard to other intestinal disorders, Hilton et al. (1997) claimed that Lb. rhamnosus GG decreased the risk of traveller's diarrhoea and Felley et al. (2001) reported that humans fed milk fermented by Lb. johnsonii exhibited a significantly reduced density of H. pylori and intensity of gastric inflammation. Treatment with a combination of Lb. acidophilus and Bif. infantis benefited neonates with necrotizing enterocolitis according to Hoyos et al. (1999), who claimed a 60% reduction in mortality due to the treatment.
Reports of the efficacy of probiotic treatment in ameliorating intestinal infection in large animals remain scarce. Zhao et al. (1998) claimed that the application of probiotic E. coli (no LAB) reduced the carriage of E. coli O157:H7 in cattle. Lema et al. (2001) observed that lambs infected with E. coli O157:H7 and then administered Lb. acidophilus displayed no beneficial effects. However, feeding the lambs a mixture of Lb. acidophilus and Streptococcus faecium, or the Streptococcus strain alone, significantly lowered numbers of the pathogenic strain. The greatest reduction in numbers was seen with the use of a mixture of Lb. acidophilus, St. faecium, Lb. casei, Lb. fermentum and Lb. plantarum. Genovese et al. reported in 2000 that an undefined competitive exclusion culture reduced the mortality and shedding of enterotoxigenic E. coli in neonatal pigs. The same group (Genovese et al., 2003) also observed that neonatal pigs treated with a similar undefined culture shed significantly lower pathogen numbers after challenge with Salmonella choleraesuis, and also exhibited reduced counts in the lower intestine. Whether or not symptoms of infection were alleviated is not described. Fedorka-Cray et al. (1999) reported that the application to Salmonella choleraesuis-challenged piglets of a competitive exclusion culture of swine origin led to reduced Salmonella counts in their faecal contents and at the ileocolic junction, as well as reduced numbers of Salmonella positive gut tissue samples. No clinical symptoms of infection were observed in any animals, including the controls.
Many studies have reported the isolation and selection of probiotic strains for use in pigs (Chang et al, 2001; Gusils et al 2002; Nemcova et al, 1997) but the results of in vivo feeding trials of animals can be variable (Simon et al, 2003). This may in part be explained by the complexity of the intestine, leading to a variation between individual animals. While undefined cultures used in competitive exclusion products can be effective in pigs, uncertainty regarding their exact composition has led to concerns that they may result in pathogen transmission. Therefore there is a need for rational selection characterisation of strains intended for use as probiotic feed additives.
Although pigs harbouring Salmonella do not generally display clinical symptoms, carriage of this pathogen in the gastrointestinal tract can leads to carcass contamination at slaughter. This may lead, in turn, to the contamination of porcine meat products. Consumption of pork products containing Salmonella leads to many cases of food borne illness in humans each year. Although certain measures have been shown to reduce the number of cases of human Salmonellosis because of pork. The economic cost associated with Salmonella infections remains high, being estimated at hundreds of millions of dollars annually to the American economy alone.
In this investigation, the pre-treatment of weaned pigs with a defined LAB culture mixture resulting in both reduced numbers of excreted Salmonella as well as an alleviation of clinical symptoms is reported. Molecular analysis of the excreted cultures indicates that the probiotic effects observed may be ascribed to two of the five cultures in the mixture.